1. Field of the Invention
This invention relates to method of and apparatus for coded moving picture data hierarchy formation and, more particularly, to method of and apparatus for coded moving picture data hierarchy formation, which may be used for moving picture transmission systems employed for television conference communication or the like under such circumstances that various different networks are interconnected.
2. Description of the Prior Art
In a prior art hierarchical coding system, low resolution picture data obtained from original moving picture, such as those in MPEG-2space scalability, are subjected to ordinary motion compensation (MC) and DCT coding, for motion prediction in the MC and DCT coding of the original moving picture.
FIG. 17 shows the prior art hierarchical coding system. One-frame picture data a is supplied to an input terminal 50. An image shrinking unit 51 shrinks the vertical and horizontal sizes of the picture data to, for instance, one half, and supplies the shrinked picture data as low resolution picture data b to a first coder. The first coder includes a subtracter 61, a DCT transformer 62, a quantizer 63, an inverse quantizer 64, an inverse DCT transformer 65, an adder 66, a frame memory 67 and a motion predictor 68. Output data from the quantizer 63 is supplied as first important data to a transmission line.
The one-frame picture data a is also supplied to a second coder. The second coder also includes a subtracter 71, a DCT transformer 72, a quantizer 73, an inverse quantizer 74, an inverse DCT transformer 75, an adder 76, a frame memory 77 and a motion predictor 78. Output data a from the quantizer 73 is supplied as second important data to the transmission line. In the second coder, for motion prediction the motion predictor 78 is supplied with data, which has been obtained by enlarging the output of the adder 66 of the first coder to the same size as the original picture with picture enlargement processer 69.
In the above prior art hierarchical coding system, the operations of the MC and DCT coding should be done twice in total, i.e., in the first and second coders. Therefore the construction and operation of the system are complicated. In addition, in the above prior art hierarchical coding system two different kinds of picture data, i.e., low and high resolution picture data b and a, should be coded. This means that the amount of picture data to be coded is increased by 25% (in case when the low resolution picture data b is one half the high resolution picture data in both the vertical and horizontal sizes), thus reducing the coding efficiency.